diff --git a/PaddleNLP/sentiment_classification/README.md b/PaddleNLP/sentiment_classification/README.md index 58af1cbbc26df254d7ddf34237183acb19829f6c..97ae00b44618bfddb2bc9320a6b7fabea7d0c420 100644 --- a/PaddleNLP/sentiment_classification/README.md +++ b/PaddleNLP/sentiment_classification/README.md @@ -29,7 +29,7 @@ 1. PaddlePaddle 安装 - 本项目依赖于 PaddlePaddle Fluid 1.3.2 及以上版本,请参考 [安装指南](http://www.paddlepaddle.org/#quick-start) 进行安装 + 本项目依赖于 PaddlePaddle Fluid 1.6 及以上版本,请参考 [安装指南](http://www.paddlepaddle.org/#quick-start) 进行安装 2. 代码安装 @@ -42,7 +42,7 @@ 3. 环境依赖 - 请参考 PaddlePaddle [安装说明](https://www.paddlepaddle.org.cn/documentation/docs/zh/1.5/beginners_guide/install/index_cn.html) 部分的内容 + Python 2 的版本要求 2.7.15+,Python 3 的版本要求 3.5.1+/3.6/3.7,其它环境请参考 PaddlePaddle [安装说明](https://www.paddlepaddle.org.cn/documentation/docs/zh/1.5/beginners_guide/install/index_cn.html) 部分的内容 ### 代码结构说明 diff --git a/PaddleNLP/sentiment_classification/reader.py b/PaddleNLP/sentiment_classification/reader.py index 778072fac81e81dcea3cda0264a15a5887532c01..fe028326e4ada9200f93c7e745fbb39274dd8b4c 100644 --- a/PaddleNLP/sentiment_classification/reader.py +++ b/PaddleNLP/sentiment_classification/reader.py @@ -72,12 +72,12 @@ class SentaProcessor(object): Generate data for train, dev or infer """ if phase == "train": - return paddle.batch(self.get_train_examples(self.data_dir, epoch, self.max_seq_len), batch_size) + return fluid.io.batch(self.get_train_examples(self.data_dir, epoch, self.max_seq_len), batch_size) #return self.get_train_examples(self.data_dir, epoch, self.max_seq_len) elif phase == "dev": - return paddle.batch(self.get_dev_examples(self.data_dir, epoch, self.max_seq_len), batch_size) + return fluid.io.batch(self.get_dev_examples(self.data_dir, epoch, self.max_seq_len), batch_size) elif phase == "infer": - return paddle.batch(self.get_test_examples(self.data_dir, epoch, self.max_seq_len), batch_size) + return fluid.io.batch(self.get_test_examples(self.data_dir, epoch, self.max_seq_len), batch_size) else: raise ValueError( "Unknown phase, which should be in ['train', 'dev', 'infer'].") diff --git a/PaddleNLP/sentiment_classification/run_classifier.py b/PaddleNLP/sentiment_classification/run_classifier.py index d16e50c8a9da3070e85a7b9103768bba477809ab..3727380128dbb7c8753c21f8dca90b0089a0cbd1 100644 --- a/PaddleNLP/sentiment_classification/run_classifier.py +++ b/PaddleNLP/sentiment_classification/run_classifier.py @@ -21,6 +21,7 @@ from nets import cnn_net from nets import bilstm_net from nets import gru_net from models.model_check import check_cuda +from models.model_check import check_version from config import PDConfig import paddle @@ -39,11 +40,11 @@ def create_model(args, """ data = fluid.layers.data( - name="src_ids", shape=[-1, args.max_seq_len, 1], dtype='int64') + name="src_ids", shape=[-1, args.max_seq_len], dtype='int64') label = fluid.layers.data( name="label", shape=[-1, 1], dtype="int64") seq_len = fluid.layers.data( - name="seq_len", shape=[-1, 1], dtype="int64") + name="seq_len", shape=[-1], dtype="int64") data_reader = fluid.io.PyReader(feed_list=[data, label, seq_len], capacity=4, iterable=False) diff --git a/PaddleNLP/sentiment_classification/run_ernie_classifier.py b/PaddleNLP/sentiment_classification/run_ernie_classifier.py index 15d5f46624b1dfa69150a7c55bff0bf7a56bfd74..0e89414d5ef85c08ba4e029b4870c77eba39f031 100644 --- a/PaddleNLP/sentiment_classification/run_ernie_classifier.py +++ b/PaddleNLP/sentiment_classification/run_ernie_classifier.py @@ -48,16 +48,21 @@ def ernie_pyreader(args, pyreader_name): labels = fluid.layers.data( name="labels", shape=[-1, 1], dtype="int64") seq_lens = fluid.layers.data( - name="seq_lens", shape=[-1, 1], dtype="int64") - - pyreader = fluid.io.PyReader(feed_list=[src_ids, sent_ids, pos_ids, input_mask, labels, seq_lens], - capacity=4, iterable=False) + name="seq_lens", shape=[-1], dtype="int64") + + pyreader = fluid.io.DataLoader.from_generator( + feed_list=[src_ids, sent_ids, pos_ids, input_mask, labels, seq_lens], + capacity=50, + iterable=False, + use_double_buffer=True) + ernie_inputs = { "src_ids": src_ids, "sent_ids": sent_ids, "pos_ids": pos_ids, "input_mask": input_mask, "seq_lens": seq_lens} + return pyreader, ernie_inputs, labels def create_model(args, @@ -299,15 +304,15 @@ def main(args): if args.do_train: train_exe = exe - train_pyreader.decorate_batch_generator(train_data_generator) + train_pyreader.set_batch_generator(train_data_generator) else: train_exe = None if args.do_val: test_exe = exe - test_pyreader.decorate_batch_generator(test_data_generator) + test_pyreader.set_batch_generator(test_data_generator) if args.do_infer: test_exe = exe - infer_pyreader.decorate_batch_generator(infer_data_generator) + infer_pyreader.set_batch_generator(infer_data_generator) if args.do_train: train_pyreader.start()